Anyone from the modern world (and a few of us from the “ancient” one of the mid- to late-1990’s) who has ever had to try and wrangle a connection out of the MacTCP screen invariably ends up exclaiming “MacTCP? MacWTF more like it!”
It may be hard to believe, but MacTCP had it’s beginnings in a simpler time…well, a less fully-featured time, anyway, before the Internet had matured and become a fixture in so many households.
MacTCP was originally designed with the sorts of TCP/IP networks which featured heavily in early academic, military, and occasionally corporate TCP networks before dialup was a thing we all tortured modems with (that is why they scream, isn’t it?).
But there are a few basic rules which makes MacTCP configuration easier, and I’ll explain them here so that you all become MacTCP experts – just in time for the rush in jobs for MacTCP engineers that’s sure to come any day now…
Manual Labour
The first rule of MacTCP is you’ll need to configure everything manually.
MacTCP knows nothing about DHCP, so you’re going to have to know three basic pieces of information.
1. The IP address you want to use for the computer running MacTCP (if you’re in emulation, this is separate to your host OS’s IP address for the purposes of MacTCP);
2. The IP address of the network router you’ll use to access the rest of the network and wider Internet; and
3. The IP address of at least one DNS server – it will usually be your router’s IP address, or it may be external to your network.
First Steps
When you open MacTCP for the first time, you’ll see something like this:
#https://blog.europlus.zone/mactcp/01.MacTCP.png#{MacTCP’s main screen}
This guide assumes you’ll be using Ethernet, as is common for devices on GlobalTalk, so click that. Unsurprisingly, the screen will now look like this:
#https://blog.europlus.zone/mactcp/02.MacTCP-a.png#{MacTCP’s main screen with Ethernet selected}
Now we get to the meat. Click “More…” and you’ll see MacTCP’s configuration screen, which should look like this:
#https://blog.europlus.zone/mactcp/03.MacTCP.png#{MacTCP’s configuration screen}
Your first task is to select Manually, instead of Server and Dynamically, in the Obtain Address section. This is what will unlock all the other settings you need to adjust:
#https://blog.europlus.zone/mactcp/04.MacTCP.png#{MacTCP’s configuration screen with Manually selected}
Ignorance is bliss
If you’ve ever played with MacTCP, the IP Address section of this window is a perplexing conundrum, wrapped inside a puzzle, etc., etc.
It also originates in the “simpler” days of the early Internet, before “classless subnets” were a thing.
Suffice it to say, you do not need to adjust anything in this section of this window!
It’s almost impossible to get anything set in here which makes sense in today’s usage of subnets and subnet masks, and you’ll likely confuse yourself, and/or muck up your settings.
So this is like an early mark - nothing to do here, move along, we’ll be finished in no time!
Gateway to Freedom
Yes, the Gateway address here is your router’s internal IP address (often 192.168.0.1 in the past, and occasionally still, but your settings will almost certainly vary from that venerable default these days).
Just enter it as you would anywhere else, and we’re two thirds done on the settings in this window, and half done overall:
#https://blog.europlus.zone/mactcp/05.MacTCP.png#{MacTCP’s configuration screen with the Gateway Address entered}
Eminent Domain
There’s an old joke about how a networking problem isn’t DNS, but the punchline is it was because every networking problem ends up being a DNS problem – well, the actual joke is better than that, sorry I’ve likely spoiled it for you if you ever see it in its original form.
The Domain Name Server Information looks much simpler than it is, but it’s not as complex as misconfiguration may make it seem.
Firstly, we need to understand that each line defines one DNS server for use by MacTCP. You get extra lines as you configure each on in turn.
These DNS servers can be internal to your network (usually your router), or they can be external ones. Internal ones tend to make troubleshooting easier in the first instance, but there’s no reason why you can’t settle on external ones down the track if that’s your preference.
Secondly, we need to understand about search domains - simply put, the Domain field in this section allows you to specify search domains. No, don’t worry, that’s not all the explanation you get.
To explain search domains, let’s think back to the early days of the Internet. You’re in your University department, or your corporate IT department, and a lot of your traffic is internal to that network, and to servers running within the network.
Now, you can imagine it would be painful to have to enter mail.someuniversity.edu and telnet.someuniversity.edu and gopher.someuniversity.edu and ftp1.someuniversity.edu and ftp2.someuniversity.edu every day when you’re in your command line (or graphical client if you’re lucky) wanting to connect to all those (or any other) internal servers.
So, the IT department would set up a DNS server for the “someuniversity.edu” domain, and that DNS server would allow you to just type “mail” or “telnet” or “ftp1” and it would append “.someuniversity.edu” (the “search domain”) to any such single word domain names it would be asked to look up - by Grabthar’s Hammer, what a savings in time and effort!
In short, the Domain field here is for such search domains. These days, it’s rare to have a DNS server set up to handle such search domains, so we’d tend to ignore that field…except…
The “search domain” for the whole internet (that is, for any domain names you enter which aren’t in otherwise-specified search domains’ DNS server/s), is represented by a single “.” character in the Domain field.
What this means for most setups in modern networks is to put that dot character in the first line’s Domain field, the DNS server’s IP address in the first line's IP Address field, and select Default on that first line – that should be all you need to do to get a working DNS configuration, and the configuration screen should now look like this (look for the dot, it’s easy to miss! The second line which has now appeared, less easy to miss, but still worth noting!):
#https://blog.europlus.zone/mactcp/06.MacTCP.png#{MacTCP’s configuration screen with all settings needed there now set}
If you look back at our list of pieces of information, there’s just one we haven’t used yet, the IP address for the machine running MacTCP. We get to set that back on the main MacTCP screen so click OK (or press Return), and we’ll be back near where we started…
#https://blog.europlus.zone/mactcp/02.MacTCP-a.png#{MacTCP’s main screen with Ethernet selected}
The Last Thing to Address
Before we finish up with the MacTCP computer’s IP address, we just need to consider the address we’re going to use, and how it relates to the rest of the network, and the DHCP range which has been specified in the router.
All DHCP servers have a range of addresses they will dole out to DHCP clients, in the simple traditional form, a router with IP address 192.168.0.1 will hand out addresses of the form 192.168.0.n where “n”: is in the DHCP range.
Each dotted number in an IP address (when expressed in decimal) can be between 0 and 255, but addresses with 0 and 255 at the end are special addresses which won’t form part of the DHCP range. The router’s IP address would normally be excluded as well, so potentially, the largest range would be 2-254 in our example network.
Many DHCP servers don’t have the range specified as that maximum, to allow for “manual” IP addresses, where all details, including the last portion of the IP address are selected, and configured, by the network administrator (on GlobalTalk, that’s you!).
Another aspect of DHCP is the ability to “reserve” addresses for specific machines. Machines are usually paired with an IP address linked to their MAC address (not “Mac”, although pronounced the same).
So a GlobalTalk administrator has two choices:
• use a manual address and be fully responsible for the choice, allocation, and ongoing avoidance of address conflicts; or
• use DHCP reservation to ensure other devices are not allocated that address, but manually enter the IP address in MacTCP.
The choice is arbitrary, but you might feel more comfortable using the DHCP reservation system as what is effectively a ledger of IP addresses, knowing the avoidance of collisions is handled by that “ledger”.
Explaining how to do the reservation is beyond the scope of this guide. For the moment, let’s assume we have selected “192.168.0.234” as the IP address for the MacTCP machine.
Whether that is DHCP-reserved or fully manually managed, enter that IP address in the MacTCP main screen, and it should look something like:
#https://blog.europlus.zone/mactcp/07.MacTCP.png#{MacTCP’s main screen with IP address entered}
That’s it – we’re done!
While your port names might change (“Ethernet", or "Ethernet Slot 2”), or you may even have a more complex configuration utilising MacIP Gateway (beyond the scope of this guide) which uses the “EtherTalk” port, or maybe you’re even going to setup “real” dialup via InterSLIP or PPP (also beyond the scope of this guide), the basic details from above will be the same, and the most important ones are in relation to ignoring the subnet settings, and getting the DNS settings right (other configurations may support Server or Dynamically as the “Obtain Address method”, but the instructions for those methods will discuss that).
I have been finding those basics have served me well in over 30 years of using MacTCP and supporting others in a variety of connection environments, and they should help GlobalTalk admins get the most out of their MacTCP-based machines – Happy Networking!
@europlus
It may be hard to believe, but MacTCP had it’s beginnings in a simpler time…well, a less fully-featured time, anyway, before the Internet had matured and become a fixture in so many households.
MacTCP was originally designed with the sorts of TCP/IP networks which featured heavily in early academic, military, and occasionally corporate TCP networks before dialup was a thing we all tortured modems with (that is why they scream, isn’t it?).
But there are a few basic rules which makes MacTCP configuration easier, and I’ll explain them here so that you all become MacTCP experts – just in time for the rush in jobs for MacTCP engineers that’s sure to come any day now…
Manual Labour
The first rule of MacTCP is you’ll need to configure everything manually.
MacTCP knows nothing about DHCP, so you’re going to have to know three basic pieces of information.
1. The IP address you want to use for the computer running MacTCP (if you’re in emulation, this is separate to your host OS’s IP address for the purposes of MacTCP);
2. The IP address of the network router you’ll use to access the rest of the network and wider Internet; and
3. The IP address of at least one DNS server – it will usually be your router’s IP address, or it may be external to your network.
First Steps
When you open MacTCP for the first time, you’ll see something like this:
#https://blog.europlus.zone/mactcp/01.MacTCP.png#{MacTCP’s main screen}
This guide assumes you’ll be using Ethernet, as is common for devices on GlobalTalk, so click that. Unsurprisingly, the screen will now look like this:
#https://blog.europlus.zone/mactcp/02.MacTCP-a.png#{MacTCP’s main screen with Ethernet selected}
Now we get to the meat. Click “More…” and you’ll see MacTCP’s configuration screen, which should look like this:
#https://blog.europlus.zone/mactcp/03.MacTCP.png#{MacTCP’s configuration screen}
Your first task is to select Manually, instead of Server and Dynamically, in the Obtain Address section. This is what will unlock all the other settings you need to adjust:
#https://blog.europlus.zone/mactcp/04.MacTCP.png#{MacTCP’s configuration screen with Manually selected}
Ignorance is bliss
If you’ve ever played with MacTCP, the IP Address section of this window is a perplexing conundrum, wrapped inside a puzzle, etc., etc.
It also originates in the “simpler” days of the early Internet, before “classless subnets” were a thing.
Suffice it to say, you do not need to adjust anything in this section of this window!
It’s almost impossible to get anything set in here which makes sense in today’s usage of subnets and subnet masks, and you’ll likely confuse yourself, and/or muck up your settings.
So this is like an early mark - nothing to do here, move along, we’ll be finished in no time!
Gateway to Freedom
Yes, the Gateway address here is your router’s internal IP address (often 192.168.0.1 in the past, and occasionally still, but your settings will almost certainly vary from that venerable default these days).
Just enter it as you would anywhere else, and we’re two thirds done on the settings in this window, and half done overall:
#https://blog.europlus.zone/mactcp/05.MacTCP.png#{MacTCP’s configuration screen with the Gateway Address entered}
Eminent Domain
There’s an old joke about how a networking problem isn’t DNS, but the punchline is it was because every networking problem ends up being a DNS problem – well, the actual joke is better than that, sorry I’ve likely spoiled it for you if you ever see it in its original form.
The Domain Name Server Information looks much simpler than it is, but it’s not as complex as misconfiguration may make it seem.
Firstly, we need to understand that each line defines one DNS server for use by MacTCP. You get extra lines as you configure each on in turn.
These DNS servers can be internal to your network (usually your router), or they can be external ones. Internal ones tend to make troubleshooting easier in the first instance, but there’s no reason why you can’t settle on external ones down the track if that’s your preference.
Secondly, we need to understand about search domains - simply put, the Domain field in this section allows you to specify search domains. No, don’t worry, that’s not all the explanation you get.
To explain search domains, let’s think back to the early days of the Internet. You’re in your University department, or your corporate IT department, and a lot of your traffic is internal to that network, and to servers running within the network.
Now, you can imagine it would be painful to have to enter mail.someuniversity.edu and telnet.someuniversity.edu and gopher.someuniversity.edu and ftp1.someuniversity.edu and ftp2.someuniversity.edu every day when you’re in your command line (or graphical client if you’re lucky) wanting to connect to all those (or any other) internal servers.
So, the IT department would set up a DNS server for the “someuniversity.edu” domain, and that DNS server would allow you to just type “mail” or “telnet” or “ftp1” and it would append “.someuniversity.edu” (the “search domain”) to any such single word domain names it would be asked to look up - by Grabthar’s Hammer, what a savings in time and effort!
In short, the Domain field here is for such search domains. These days, it’s rare to have a DNS server set up to handle such search domains, so we’d tend to ignore that field…except…
The “search domain” for the whole internet (that is, for any domain names you enter which aren’t in otherwise-specified search domains’ DNS server/s), is represented by a single “.” character in the Domain field.
What this means for most setups in modern networks is to put that dot character in the first line’s Domain field, the DNS server’s IP address in the first line's IP Address field, and select Default on that first line – that should be all you need to do to get a working DNS configuration, and the configuration screen should now look like this (look for the dot, it’s easy to miss! The second line which has now appeared, less easy to miss, but still worth noting!):
#https://blog.europlus.zone/mactcp/06.MacTCP.png#{MacTCP’s configuration screen with all settings needed there now set}
If you look back at our list of pieces of information, there’s just one we haven’t used yet, the IP address for the machine running MacTCP. We get to set that back on the main MacTCP screen so click OK (or press Return), and we’ll be back near where we started…
#https://blog.europlus.zone/mactcp/02.MacTCP-a.png#{MacTCP’s main screen with Ethernet selected}
The Last Thing to Address
Before we finish up with the MacTCP computer’s IP address, we just need to consider the address we’re going to use, and how it relates to the rest of the network, and the DHCP range which has been specified in the router.
All DHCP servers have a range of addresses they will dole out to DHCP clients, in the simple traditional form, a router with IP address 192.168.0.1 will hand out addresses of the form 192.168.0.n where “n”: is in the DHCP range.
Each dotted number in an IP address (when expressed in decimal) can be between 0 and 255, but addresses with 0 and 255 at the end are special addresses which won’t form part of the DHCP range. The router’s IP address would normally be excluded as well, so potentially, the largest range would be 2-254 in our example network.
Many DHCP servers don’t have the range specified as that maximum, to allow for “manual” IP addresses, where all details, including the last portion of the IP address are selected, and configured, by the network administrator (on GlobalTalk, that’s you!).
Another aspect of DHCP is the ability to “reserve” addresses for specific machines. Machines are usually paired with an IP address linked to their MAC address (not “Mac”, although pronounced the same).
So a GlobalTalk administrator has two choices:
• use a manual address and be fully responsible for the choice, allocation, and ongoing avoidance of address conflicts; or
• use DHCP reservation to ensure other devices are not allocated that address, but manually enter the IP address in MacTCP.
The choice is arbitrary, but you might feel more comfortable using the DHCP reservation system as what is effectively a ledger of IP addresses, knowing the avoidance of collisions is handled by that “ledger”.
Explaining how to do the reservation is beyond the scope of this guide. For the moment, let’s assume we have selected “192.168.0.234” as the IP address for the MacTCP machine.
Whether that is DHCP-reserved or fully manually managed, enter that IP address in the MacTCP main screen, and it should look something like:
#https://blog.europlus.zone/mactcp/07.MacTCP.png#{MacTCP’s main screen with IP address entered}
That’s it – we’re done!
While your port names might change (“Ethernet", or "Ethernet Slot 2”), or you may even have a more complex configuration utilising MacIP Gateway (beyond the scope of this guide) which uses the “EtherTalk” port, or maybe you’re even going to setup “real” dialup via InterSLIP or PPP (also beyond the scope of this guide), the basic details from above will be the same, and the most important ones are in relation to ignoring the subnet settings, and getting the DNS settings right (other configurations may support Server or Dynamically as the “Obtain Address method”, but the instructions for those methods will discuss that).
I have been finding those basics have served me well in over 30 years of using MacTCP and supporting others in a variety of connection environments, and they should help GlobalTalk admins get the most out of their MacTCP-based machines – Happy Networking!